11:10
Session 7D: Simulation methods & control
Chair: Hua Tian
11:10
20 mins
|
Non-linear State Estimator for Advanced Control of an ORC Test Rig for Geothermal Application
Roberto Pili, Sebastian Eyerer, Fabian Dawo, Christoph Wieland, Hartmut Spliethoff
Abstract: Organic Rankine Cycle Systems (ORC) are able to convert efficiently low-temperature geothermal heat sources into mechanical and electrical power or combined heat and power. Especially when producing both heat and power, high operational flexibility is necessary to meet the heat demand and supply electricity in an efficient way. Advanced controllers, as linear quadratic integrators, can be used to guarantee the required flexibility of the ORC unit. Such advanced controllers rely on information on the system state, which is in general non-fully measurable. To reach this goal, state estimators are used and analyzed in this work. First, a dynamic model of an ORC test rig for geothermal application is developed and validated against experimental data. Subsequently, a non-linear state estimator for the ORC evaporator coupled with a screw expander is designed and tested on a benchmark case. The considered estimator is an Unscented Kalman Filter based on a finite volume model of the evaporator. The results show a good agreement among the dynamic and observer model and the experimental data. The estimated states by the filter can be fed to advanced single- or multi-variable controllers to maximize the ORC net power output and revenues. In the next future, the integration between the observer and an advanced control system will be analyzed and the performance under changes in electric load and heat demand will be tested both via simulations and experiments on the test rig.
|
11:30
20 mins
|
Artificial Neural Networks for Real-time Model Predictive Control of Organic Rankine Cycles for Waste Heat Recovery
Yannic Vaupel, Adrian Caspari, Nils C. Hamacher, Wolfgang R. Huster, Adel Mhamdi, Ioannis G. Kevrekidis, Alexander Mitsos
Abstract: Recovering waste heat from the exhaust gas of heavy-duty diesel trucks using a bottoming organic Rankine cycle (ORC) is a promising option to reduce fuel consumption. In contrast to most other applications of ORCs, e.g., geothermal or solar-thermal, the heat source in automotive applications is subject to strong fluctuations with limited predictability. Consequently, controlling the ORC system to maintain safe and efficient operation is a challenging task. Nonlinear model predictive control (NMPC) has been proposed for ORC systems and showed promising in silico results. It suffers, however, from a high computational expense and real-time capable implementation on vehicle hardware is questionable. Several methods are available that aim at shifting the majority of the computational load to the design phase of the controller, reducing on-line resource demand. In this work, we apply artificial neural networks (ANN) in silico to learn the control law of the NMPC controller off-line. We obtain training data from various NMPC scenarios with different initial conditions and heat source conditions using our in-house dynamic optimization tool DyOS. Subsequently, we apply the ANN-based controller in silico to different scenarios of transient heat source conditions. We compare the results to the NMPC solution obtained with DyOS and findings indicate that performance loss associated with the ANN-based controller is marginal while the control policy can be obtained at negligible computational cost.
|
11:50
20 mins
|
Toward a High Resolution Real Gas Finite Volume Solver with Multi Optimal Order Detection
Michael Deligant, Xesus Nogueira, Sofiane Khelladi, Emilie Sauret, Brian Reding
Abstract: The accurate predictions of the ORC expander performance rely on validated numerical tools that take into account the full complexity of the underlying physics. The expansion of organic vapor in turbomachines rotor and stator features non-ideal gas behavior with chocked flow in transonic conditions and supersonic expansion. In this paper, a finite volume solver using moving least squares approximations for higher order reconstruction is used. The real gas properties are taken into account using lookup tables with Tabular Taylor Series Expansion. The SLAU approximate Riemann solver is used for its compatibility with real gas computations. The conventional/classical slope limiter approach to handle shocks is replaced with the a posteriori paradigm for the local order reduction (MOOD). The developments are validated by comparing available solution of supersonic expansion of R245fa in a converging diverging nozzle test case.%ideal gas then real
Further developments will focus on the CFD of turbulent non ideal dense gas expansions with implicit large eddy scale techniques using MOOD and automatic dissipation adjustment (ADA) for turbomachinery applications.
|
12:10
20 mins
|
Influence of Superheated Vapor in Binary Cycle with Working Fluid R123 Utilizing Low-temperature Geothermal Resources
Denny Surinda, Wahyu Caesarendra, Totok Prasetyo, Taufik Taufik
Abstract: The binary cycle on Organic Rankine Cycle (ORC) system has been appropriated technology for generating electricity in order to utilize low-temperature geothermal resources. The degree of superheated vapor attracts attention to be studied further because it is the last point to absorb heat energy from geothermal heat sources and influence the amount of expansion power produced by the expander. Therefore, achieving high ORC system efficiency requires a parameter of superheated vapor degree. This work presents an experimental study on binary cycle applying R123 as working fluid to investigate the effect a variety of superheated vapor degree on the ORC efficiency. Geothermal heat sources are simulated with lubricant oil as an external heat source to provide input heat to ORC system. Temperature High inlet (TH in) evaporator is made to remain at 120°C during the experiment, while mass flow rate is adjusted to make superheated vapor variations, namely set at 5°C, 7°C, 9°C, 11°C, and 13°C. Furthermore, the effect is observed on heat transfer inlet, pinch, heat transfer coefficient, expander work output, isentropic efficiency, expander shaft power, power generation, thermal efficiency, and ORC efficiency. The experimental results showed that mass flow rate nearly remains unchanged on variety of superheated vapor. The range of heat transfer inlet, pinch temperature and heat transfer coefficient are 27.89 kJ/kg-25.34 kJ/kg, 9.35°C-4.08°C, 200.62 W/m2 .°K-232.54 W/m2 .°K, respectively. The expander work output demonstrates a slight decrease with range 2.56 kJ/s-1.98 kJ/s, while generator produces relatively a same trend electricity with range 1.37 kW-1.37 kW. The maximum back work ratio and ORC efficiency are 0.079 at superheated of 13°C and 8.6% at superheated of 3°C, respectively. In conclusion, ORC system efficiency can be triggered by many parameters, including the amount of temperature on the exit side of evaporator. The superheated vapor of working fluid R123 to higher temperature has affected a decrease in ORC system efficiency due to the decrease in heat transfer inlets, although based on the theory reveals that the work total has increased. Further investigation have found that the magnitude of the mass flow rate affects the behavior of component ORC system.
|
|